EpipwR: efficient power analysis for EWAS with continuous outcomes

Area of Science:

• Genomics
• Biostatistics
• Computational Biology

Background:

• Epigenome-wide association studies (EWAS) are crucial for understanding complex diseases.
• Existing power estimation tools for EWAS are limited, especially for non-case-control designs.
• There is a need for user-friendly tools that support diverse EWAS study designs.

Purpose of the Study:

• To introduce EpipwR, an open-source R-package for efficient EWAS power estimation.
• To expand power calculation capabilities beyond traditional case-control studies.
• To provide researchers with a tool for planning more accurate and comprehensive EWAS.

Main Methods:

• EpipwR utilizes a quasi-simulated approach for power estimation.
• It generates data for relevant CpG sites and calculates P-values directly for non-associated sites.
• The package leverages empirical EWAS datasets to guide data generation.

Main Results:

• EpipwR demonstrates efficient power estimation for EWAS with continuous or binary outcomes.
• Numerical studies confirm the influence of empirical datasets on correlation and power.
• EpipwR outperforms existing power calculation alternatives on both simulated and real EWAS data.

Conclusions:

• EpipwR provides a valuable and accurate tool for EWAS power analysis.
• The R-package supports a wider range of study designs than previously available.
• EpipwR is accessible on Bioconductor and GitHub, facilitating its adoption by researchers.